DRAFT VERSION APRIL 25, 2012 Preprint typeset using LATEX style emulateapj v. 8/13/10

THE DISCOVERY OF AN ULTRA-FAINT STAR CLUSTER IN THE CONSTELLATION OF URSA MINOR1

R. R.MUÑOZ2,3 ,M.GEHA2 , P. CÔTÉ4,L.C.VARGAS2,F.A.SANTANA3, P.STETSON4,J.D.SIMON5 &S.G.DJORGOVSKI6,7 Draft version April 25, 2012

ABSTRACT We report the discovery of a new ultra-faint globular cluster in the constellation of Ursa Minor, based on stellar photometry from the MegaCam imager at the Canada-France-Hawaii Telescope (CFHT). We find that this cluster, Muñoz 1, is located at a distance of 45 ± 5kpc and at a projected distance of only 45′ from the center of the Ursa Minor dSph galaxy. Using a Maximum Likelihood technique we measure a half-light radius of 0′.5, or equivalently 7 pc and an ellipticity consistent with being zero. We estimate its absolute magnitude − ± +160 to be MV = 0.4 0.9, which corresponds to LV = 120−65 L⊙ and we measure a heliocentric radial velocity of −137 ± 4kms−1 based on Keck/DEIMOS spectroscopy. This new satellite is separate from Ursa Minor by ∼ 30kpcand110kms−1 suggesting the cluster is not obviously associated with the dSph, despite the very close angular separation. Based on its photometric properties and structural parameters we conclude that Muñoz 1 is a new ultra-faint stellar cluster. Along with Segue 3 this is one of the faintest stellar clusters known to date. Subject headings: galaxies: photometry - globular clusters: general - Galaxy: halo - Local Group

1. INTRODUCTION TABLE 1 Over the last seven years, and thanks to the advent of the PARAMETERSOF MUÑOZ 1 Sloan Digital Sky Survey (SDSS; York et al. 2000), seventeen new Milky Way satellites have been discovered (e.g., Willman Parameter Mean Uncertainty et al. 2005;Zuckeret al. 2006a,b;Belokurovet al. 2006,2007; ′′ Irwin et al. 2007). The majority of these objects correspond to α0,exp (h m s) 15:01:48.02 ±13 ± ′′ a new class of ultra low-luminosity dwarf galaxies, based on δ0,exp (d m s) +66:58:07.3 8 ± their kinematic and metallicity properties (e.g., Muñoz et al. rh,exp (arcmin) 0.49 0.19 rh,exp (pc) 7.1 ±2.8 2006; Martin et al. 2007; Simon & Geha 2007), with lumi- ± ∼ − rh,P (arcmin) 0.49 0.15 nosities ranging from MV 8 for the brighter of these ob- rh,P (pc) 7.1 ±2.1 ± jects down to an extreme MV ∼ −1.5 for the faintest (Martin rc (arcmin) 15" 9” ± et al. 2008). Given their extremely low stellar contents, they rc (pc) 4.5 1.9 rt (arcmin) 3.6 ±2.2 are commonly referred to as the Ultra-Faint Dwarfs (UFDs). rt (pc) 65 ±28 A few other systems are clearly low luminosity globular clus- MV (Chabrier) −0.4 ±0.9 + − ters, which include Koposov 1 and 2 (MV = −1 and −2 respec- LV (L⊙) 120 160, 65 −2 + − tively; Koposov et al. 2007) and the extreme case of Segue 3 µ0,V (mag arcsec ) 26.3 1.6, 2.1 −1 ∼ vr (km s ) −137 ±4 (MV 0.0; Belokurov et al. 2010; Fadely et al. 2011). ± Despite the enormous success of satellites searches in d (kpc) 45 5 SDSS, this survey covers only about a fourth of the sky down to a magnitude limit of r ∼ 22.5. In practice this means systems. that the faintest objects can only be detected out to distances In this Letter we present the discovery of a new extremely smaller than a few tens of kiloparsecs (see Fig. 9 in Tollerud low-luminosity, outer halo stellar cluster, in the outskirts of et al. 2008), and thus the majority of the Milky Way’s virial the Ursa Minor dwarf spheroidal (dSph) galaxy. This discov- volume remains unsearched for UFDs and other faint stellar ery was made just outside the SDSS footprint using signifi- cantly deeper photometry which allowed us to probe farther 1 Based on observations obtained at the Canada-France-Hawaii Telescope into the Milky Way halo. In §2 we detail the data-set and the (CFHT) which is operated by the National Research Council of Canada, discovery. In §3 we present the results from the photometric the Institut National des Sciences de l’Univers of the Centre National de la analysis and in §4 we show radial velocity data recently ob- Recherche Scientifique of France, and the University of Hawaii. Spectro- scopic data presented herein were obtained at the W. M. Keck Observatory, tained. Finally, in §5 we discuss and summarize our results. which is operated as a scientific partnership among the California Institute of Technology, the University of California and the National Aeronautics and 2. DATA AND DISCOVERY Space Administration. The discovery of this new system was made serendipi- 2 Astronomy Department, Yale University, New Haven, CT 06520, USA 3 Departamento de Astronomía, Universidad de Chile, Camino El Obser- tously while analyzing photometric data for the Ursa Minor vatorio 1515, Las Condes, Santiago, Chile ([email protected]) dSph galaxy taken with the MegaCam imager at the Canada- 4 Herzberg Institute of Astrophysics, National Research Council of France-Hawaii Telescope (CFHT). MegaCam is a wide-field Canada, Victoria, BC, V9E 2E7, Canada × 5 imager consisting of 36 2048 4612 pixel CCDs, covering Observatories of the Carnegie Institution of Washington, 813 Santa Bar- almost a full 1 × 1deg2 field of view with a pixel scale of bara St., Pasadena, CA 91101, USA − 6 Astronomy Department, California Institute of Technology, Pasadena, 0”.187 pixel 1. These data were taken as part of a larger pro- CA, 91125, USA gram aimed at obtaining deep wide-field imaging of all bound 7 Distinguished Visiting Professor, King Abdulaziz University, Jeddah, Saudi Arabia. stellar over-densities in the Milky Way halo beyond 25 kpc (R. R. Muñoz et al. 2012, in preparation). Ursa Minor was ob- 2 Muñoz et al.

structural parameters of the object, i.e., scale length (half-light radius or King core and tidal radii depending on the profile), coordinates of the center of the cluster, ellipticity, position an- gle and background density. To estimate parameter uncertain- ties we carry out a bootstrap analysis of 10000 realizations of the photometric data. Table 1 shows the resulting parameters. We obtain a half- light radius of 7.1 ± 2.1pc using both an exponential and Plummer profile, whereas a King profile yields rcore =4.5 ± 1.9pc and rtidal = 65 ± 28pc. Figure 3 shows the number density profile for Muñoz 1. We have overplotted the three best-fit profiles, all of which give a reasonable description of the light distribution. The right panel of this figure shows a density contour map of the ob- ject, where the contours represent 2, 3, 5, 8, 12 and 18 − σ levels above the background density. These contours show FIG.1.— 1′.2 × 1′.2, r−band view of Muñoz 1. The image is roughly no evident elongation or tidal features. Applying a bootstrap centered on the cluster. North is up, east is to the left. analysis where we generate 10000 realizations of the photo- served on the nights of UT June 12-14 and July 07-13, 2010 metric data (Walsh et al. 2008; Muñoz et al. 2010), we deter- under dark conditions with typical seeing of 0.7 − 0”.9. Four mine that the asymmetries observed in the outer parts, toward different, slightly overlapping fields were observed for a total the south and west directions, are not statistically significant area coverage of nearly 2 × 2deg2. In each field, the center of due to the very low number of stars. the dSph was placed in one of the corners so that, when com- In addition to the structural parameters, we estimate the ab- bined, the galaxy is located at the center of the covered area. solute magnitude of the new object. We follow the method We obtained six dithered exposures of 360 seconds in both described in Muñoz et al. (2010)which relies in the total num- the g− and r−band. A standard dithering pattern was used to ber of stars that belong to the cluster and not on the sum of cover both small and large gaps present between the chips. their fluxes. As shown by Martin et al. (2008) the low num- MegaCam images are deliveredto the user pre-processedby ber of stars identified in the ultra-faint systems, especially the CFHT team using the “Elixir" package (Magnier & Cuil- true in the case of Muñoz 1, makes traditional methods of landre 2004). Subsequent photometric measurements were adding the individual fluxes of the member stars very sen- carried out using first DAOPHOT/Allstar, and later running sitive to the inclusion (or exclusion) of potential members, the ALLFRAME package on the processed frames, following especially at brighter magnitudes. To alleviate shot noise is- the procedure outlined in Stetson (1994). Finally, astrometric sues we use an alternative method based on a model stellar solutions were calculated using the freely available SCAMP8 population. For this we use a theoretical luminosity function code, and photometric calibration was carried out by direct that best describes the photometric properties of the cluster, in comparison with data from the SDSS-Data Release 7 (DR7; this case a 12.5Gyr old population with [Fe/H]= −1.5 (Dotter Abazajian & Sloan Digital Sky Survey 2009). For more de- et al. 2008). We then integrate the LF to obtain the total flux tails on the method see Muñoz et al. (2010). down to a given magnitude limit. The last step is to scale this As seen in Figure 1 and the left panel of Figure 2, we dis- flux using the total number of stars that belong to the cluster covered a centrally concentrated over-density of stars 45′ to down to the same magnitude limit. We estimate the corre- the south-west of Ursa Minor. Based on the structural pa- spondinguncertaintythrough a bootstrap analysis. This yields rameters determined below, we presume this object to be an MV = −0.4 ± 0.9mag assuming a Chabrier initial mass func- +160 ultra-faint globular cluster and therefore name it Muñoz 1. In tion (Chabrier 2001), which corresponds to LV = 120−65 L⊙. §5 we discuss in more depth our reasoning for this classi- Only one old star cluster has a published luminosity lower fication. Figure 2 also shows the color-magnitude diagram than Muñoz 1, the Segue 3 stellar cluster with a total lumi- (CMD) of the central region of Muñoz 1 where a distinct nosity of MV = −0.0 ± 0.8mag (Fadely et al. 2011), although main-sequence turn-off is observed. Assuming the stars in given the large uncertainties in both measurements it is un- the red-giant branch (RGB) region are members of the cluster clear which cluster is actually fainter. we fit an isochrone at a distance of 45kpc, an age of 12.5Gyr 4. SPECTROSCOPIC PROPERTIES and [Fe/H]= −1.5. We note that this fit is very tentative given the very low number of stars that belong to the system. The Spectroscopic data of photometrically selected Muñoz 1 right panel of the figure compares the CMDs of UMi and the candidate stars were taken on UT May 28, 2011 with the new object showing a clear difference in distance modulus. Keck II 10-m telescope and the DEIMOS spectrograph (Faber et al. 2003). One Keck/DEIMOS multi-slit mask was ob- 3. STRUCTURAL PROPERTIES served with the 1200 line mm−1 grating covering a wavelength We carry out a maximum-likelihood analysis of the photo- region 6400−9100Å with a spectral dispersion of 0.33Å. The metric data for Muñoz 1 following the method of Martin et al. mask was observed for 7800seconds under relatively poor (2008)as described in Muñoz et al. (2010). This method starts seeing conditions which varied between 1.5 − 2′′. Spectra by assuming a shape for the underlying light distribution; in were reduced using a modified version of the spec2d soft- this case we have tried three of the most commonly used pro- ware pipeline (version 1.1.4) developed by the DEEP2 team, files for UFDs, a King (King 1962), exponential and Plummer and radial velocities were determined using the software de- (Plummer 1911) density laws. We then fit simultaneously the scribed by Simon & Geha (2007). For additional details, we refer the reader to Geha et al. (2009). 8 See http:/www.astromatic.net/software/scamp. Radial velocities were successfully measured for 24 of the A new satellite in Ursa Minor 3

′ FIG. 2.— left: Star count map of the Ursa Minor dSph field. The primary object in this field is the dSph galaxy. Muñoz 1 lies 45 away from the center of the field in the south-west direction, near (−1.8◦,−0.25◦). middle: g vs g − r CMD of stars in the Muñoz 1 region (within 2′ of the measured center). The best fit isochrone is overplotted, corresponding to a Dartmouth (Dotter et al. 2008), 12.5 Gyr, [Fe/H]= −1.5 at a distance of ∼ 45 kpc using a reddening value of E(g − r) = 0.027, adopted from the Schlegel et al. (1998) maps. right: CMD of Ursa Minor stars (small black points) overplotted on Muñoz 1 stars (large red symbols). Red and blue solid lines represent best fit isochrones for both the Ursa Minor dSph and Muñoz 1 respectively. 47 extracted spectra. Stars for which we could not measure a matching an observed Keck/DEIMOS spectrum to a grid of velocity were primarily very low S/N spectra of faint objects models with different Teff and [Fe/H]. This is in contrast to near the expected main sequence turn-off of Muñoz 1 (r ∼ most high resolution abundances, which are based on equiva- 22 − 22.5). lent widths and require resolved, unblended lines. Our spec- The resulting velocity distribution is shown in Figure 4, tral resolution is not sufficient for such type of analysis. We along with the correspondingspatial and color-magnitude dis- note, however, that our well-characterized noise array enables tributions. While the velocity histogram does not show a clear us to calculate realistic error bars. significant peak that could be interpreted as the velocity signa- We measure metallicities for two stars kinematically asso- ture of Muñoz 1, if we restrict ourselves to the most centrally ciated with Ursa Minor and one star associated with Muñoz 1. concentrated stars, what appears as a cold peak is indeed dis- The two Ursa Minor stars have metallicities [Fe/H]= −2.78 ± cerned. Using the four stars closest to the derived center of 0.60 and −2.30 ± 0.68, while the Muñoz 1 star has [Fe/H]= the cluster, plus a fifth star possibly associated with it based −1.46±0.32. These results are consistent with the isochrone- on its radial velocity, distance to the cluster and position in derived metallicity. Due to the paucity of Fe II absorption the CMD, we estimate a mean velocity of −137 ± 4km s−1. lines in cool RGB spectra, we are unable to calculate sur- Given the small number of stars and large velocity uncertain- face gravity from the spectroscopic data. Hence, to esti- ties, we estimate only a one sigma upper limit on the velocity mate [Fe/H] we had to fix log g from the best-fit isochrone. − dispersion of σ< 4.7km s 1. Given the size and luminosity The code also outputs a value for Teff, which can be com- of Muñoz 1, we predict its velocity dispersion to be between pared to the value obtained using the de-reddened CFHT 0.2 − 0.3km s−1 based on Wolf et al. (2010) and a mass-to- photometry. For the case of the cluster star, we obtain a ± light ratio consistent with old stars only. Our measured ve- Teff = 5160 320K, in good agreement with the photometric locity dispersion is consistent with this value, but we cannot value of Teff = 5290K. rule out a scenario where the mass of Muñoz 1 is significantly larger than its stellar mass. 5. SUMMARY AND DISCUSSION Two other stars with radial velocities in the vicinity of this We report the discovery of a new ultra-faint stellar sys- peak do not live near the CMD of Muñoz 1, nor are they close tem, Muñoz 1, in the Ursa Minor constellation, located only to the center of the object, and thereforewe consider them less 45′ from the center of the Ursa Minor dSph. The discovery likely to be associated with the cluster. We also observe three was made serendipitously while analyzing photometric data stars consistent with the published radial velocity of Ursa Mi- for Ursa Minor, taken with the wide-field MegaCam imager nor of −245 km s−1 (Muñoz et al. 2005). These stars are on the CFHT. We visually fit a 12.5Gyr isochrone from the less spatially concentrated than the Muñoz 1 stars and yield Dartmouth database (Dotter et al. 2008), with a metallicity a mean velocity of −253 ± 6km s−1, in good agreement with of [Fe/H]= −1.5 at a distance of ∼ 45kpc. While the close the observed value for the dSph galaxy. angular distance between these two Galactic satellites might A subset of stars in our Keck/DEIMOS kinematic sam- suggest an association, we deem it unlikely. We have esti- −1 ple have sufficient S/N (at least 10Å ) to determine their mated the escape velocity from Ursa Minor at the projected metallicities. We determine [Fe/H] based on fitting synthetic distance of Muñoz 1, i.e., 1kpc, assuming the dSph is a point × 8 ∼ spectra, as described in Kirby et al. (2008). This medium- mass of 1 10 M⊙ (Wolf et al. 2010). We obtain vesc 30 −1 ∼ −1 resolution technique allows us to calculate Teff and [Fe/H] by km s , much smaller than the 110km s of radial velocity difference. Moreover, at 30kpc, the distance separating the 4 Muñoz et al.

FIG. 3.— Left: Number density profile for Muñoz 1. Plummer, King and exponential profiles have been fitted to the data using a maximum likelihood method. The dot-dashed line represents the measured background density. Right: Iso-density contours for Muñoz 1. Contours level correspond to 2,3,5,8,12 and 18−σ level over the background density. cluster from the dSph along the line of sight, vesc decreases galaxy, without a reliable measurement of its velocity disper- down to only a few km s−1. sion of metallicity spread, we view it as much more likely that Using a maximum-likelihood technique, following the it is a globular cluster. The luminosity derived for this object, method of Martin et al. (2008), we have simultaneously de- makes it one of the faintest globular cluster discovered, only rived structural parameters along with background density nominally surpassed by Segue 3, which has an absolute mag- for Muñoz 1. Assuming different density profiles we have nitude of MV =0.0 ± 0.8 (Fadely et al. 2011). Given the large derived consistent half light radii of rh,exp = 7.1 ± 2.8pc, uncertainties of both measurements, it is unclear which one rh,P =7.1 ± 2.1pc for exponential and Plummer profiles re- of them is indeed fainter. The density profile of this object is spectively, and a King core and tidal radii of rc =4.5±1.9and well fitted by typical cluster density profiles showing no ob- rt = 65 ± 28pc respectively. The total luminosity of Muñoz 1, vious hints for significant tidal disruption. on the other hand, was derived using the method described The discovery of this cluster was made in the context of a in Muñoz et al. (2010), which relies in the total number of larger program aimed at obtaining deep and wide-field pho- stars in the object instead of their individual fluxes, obtain- tometry for all outer halo satellites regardless of morpho- ing an absolute magnitude of MV = −0.4 ± 0.9, equivalent logical classification. At the time of this publication, we +160 2 to LV = 120−65 L⊙. Based on its photometric properties, we have imaged roughly 40deg down to a magnitude limit of argue that this new Galactic satellite corresponds to an ex- r ∼ 24.5 − 25 depending on the object. As pointed out by tremely faint globular cluster. Since the discovery of the Tollerud et al. (2008), surveys deeper than SDSS are bound to UFDs, there has been some debate in the literature about the uncover fainter and more distant ultra-faint systems. Naively classification of stellar over-densities based solely on photo- extrapolating the discovery of at least one ultra-faint cluster in metric information. It has been argued, for instance, that ob- our surveyed area to the entire sky, yields potentially a thou- jects like Segue 1 and Coma Berenices should be considered sand similar objects still uncovered. While this number is large globular clusters based on the assumption that galax- not intended as a firm prediction given its large uncertainty, ies cannot have scale lengths smaller than ∼ 100pc (Gilmore it foreshadows a much more complex picture of the globular et al. 2007). On the other hand, spectroscopic information, cluster system in the outer halo than we have today. i.e., velocity dispersions and metallicities, have been used to We acknowledge Robert Zinn for useful discussions. This argue that despite their sizes, these systems correspond to dark work was supported in part by the facilities and staff of the matter dominated galaxies (Simon & Geha 2007; Geha et al. Yale University Faculty of Arts and Sciences High Perfor- 2009). For Muñoz 1, the spectroscopic information is in- mance Computing Center. R.R.M. acknowledges support sufficient to determine unambiguously whether the object is from the GEMINI-CONICYT Fund, allocated to the project dark matter dominated or not. However, this satellite is sig- N◦32080010, from CONICYT through project BASAL PFB- nificantly smaller than any of the known dwarf galaxies, in- 06 and the Fondo Nacional de Investigación Científica y Tec- cluding the UFDs, it lies within the size distribution of other nológica (Fondecyt project N◦1120013). M. G. acknowl- Milky Way globular clusters with structural parameters simi- edges support from the National Science Foundation under lar to other outer halo clusters like Palomar 13 (Bradford et al. award number AST-0908752 and the Alfred P. Sloan Founda- 2011), and while in principle it could be the smallest known tion. S.G.D. acknowledgespartial support from the NSF grant A new satellite in Ursa Minor 5

FIG. 4.— Keck/DEIMOS spectroscopic coverage of Muñoz 1, showing the spatial distribution (left), color-magnitude diagram (middle) and velocity distribution of targets (right). Red symbols in each panel indicate stars possibly associated with Muñoz 1, while blue symbols are stars likely associated with Ursa Minor. AST-0909182.

REFERENCES Abazajian K., Sloan Digital Sky Survey f. t., 2009, ApJS, 182, 543 Magnier E. A., Cuillandre J.-C., 2004, PASP, 116, 449 Belokurov V., et al., 2006, ApJ, 647, L111 Martin N. F., de Jong J. T. A., Rix H.-W., 2008, ApJ, 684, 1075 Belokurov V., et al., 2007, ApJ, 654, 897 Martin N. F., Ibata R. A., Chapman S. C., Irwin M., Lewis G. F., 2007, Belokurov V., Walker M. G., Evans N. W., Gilmore G., Irwin M. J., Just D., MNRAS, 380, 281 Koposov S., Mateo M., Olszewski E., Watkins L., Wyrzykowski L., 2010, Muñoz R. R., Carlin J. L., Frinchaboy P. M., Nidever D. L., Majewski S. R., ApJ, 712, L103 Patterson R. J., 2006, ApJ, 650, L51 Bradford J. D., Geha M., Muñoz R. R., Santana F. A., Simon J. D., Côté P., Muñoz R. R., Frinchaboy P. M., Majewski S. R., Kuhn J. R., Chou M.-Y., Stetson P. B., Kirby E., Djorgovski S. G., 2011, ApJ, 743, 167 Palma C., Sohn S. T., Patterson R. J., Siegel M. H., 2005, ApJ, 631, L137 Chabrier G., 2001, ApJ, 554, 1274 Muñoz R. R., Geha M., Willman B., 2010, AJ, 140, 138 Dotter A., Chaboyer B., Jevremovic´ D., Kostov V., Baron E., Ferguson Plummer H. C., 1911, MNRAS, 71, 460 J. W., 2008, ApJS, 178, 89 Schlegel D. J., Finkbeiner D. P., Davis M., 1998, ApJ, 500, 525 Faber S. M., et al., 2003, in Instrument Design and Performance for Simon J. D., Geha M., 2007, ApJ, 670, 313 Optical/Infrared Ground-based Telescopes. Edited by Iye & Moorwood, Stetson P. B., 1994, PASP, 106, 250 Proceedings of the SPIE, Volume 4841, pp. 1657 The DEIMOS Tollerud E. J., Bullock J. S., Strigari L. E., Willman B., 2008, ApJ, 688, 277 spectrograph for the Keck II Telescope: integration and testing Walsh S. M., Willman B., Sand D., Harris J., Seth A., Zaritsky D., Jerjen H., Fadely R., Willman B., Geha M., Walsh S., Muñoz R. R., Jerjen H., Vargas 2008, ApJ, 688, 245 L. C., Da Costa G. S., 2011, AJ, 142, 88 Willman B., et al., 2005, ApJ, 626, L85 Geha M., Willman B., Simon J. D., Strigari L. E., Kirby E. N., Law D. R., Wolf J., Martinez G. D., Bullock J. S., Kaplinghat M., Geha M., Muñoz Strader J., 2009, ApJ, 692, 1464 R. R., Simon J. D., Avedo F. F., 2010, MNRAS, 406, 1220 Gilmore G., Wilkinson M. I., Wyse R. F. G., Kleyna J. T., Koch A., Evans York D., et al., 2000, AJ, 120, 1579 N. W., Grebel E. K., 2007, ApJ, 663, 948 Zucker D. B., et al., 2006a, ApJ, 650, L41 Irwin M., et al., 2007, ApJ, 656, L13 Zucker D. B., et al., 2006b, ApJ, 643, L103 King I., 1962, AJ, 67, 471 Kirby E. N., Guhathakurta P., Sneden C., 2008, ApJ, 682, 1217 Koposov S., de Jong J. T. A., Belokurov V., Rix H.-W., Zucker D. B., Evans N. W., Gilmore G., Irwin M. J., Bell E. F., 2007, ApJ, 669, 337